Iron is a widespread structural material in nuclear applications. However, some benchmark studies show that for many libraries, including TENDL-2019, producing good iron evaluations is still
an open challenge. This work aims at improving TENDL-2019⁵⁶Fe and ⁵⁴Fe evaluations through feedbacks from experimental data and Monte Carlo results of the iron benchmark ASPIS. Part of
the T6 code infrastructure was used to produce the evaluations. In the new evaluations, particular attention to resonance parameters and formalism was paid to address the importance of elastic scattering. The total inelastic scattering and inelastic continuum cross sections were shaped by modifying the input parameters of TALYS, one of the T6 codes. Additionally, it was necessary to normalise some reaction channels to those of JENDL-4.0. With the new files ASPIS results, obtained with the Monte Carlo code TRIPOLI-4®, improved noticeably compared to TENDL-2019. The evaluations and parameter setting modifications are to be included into TENDL-2021.